Re 1-3) Most discussions can actually be distilled to a single pointed issue or a very specific discussion (https://www.kialo.com/should-there-be-a-universal-basic-inco...).
You can then ask in a follow up, using a multi thesis discussion, which variant of e.g. legislation is the best (https://www.kialo.com/what-is-the-best-drug-regulation-syste...).
Or you can start with a multi and discuss various options, as e.g. done here: (https://www.kialo.com/who-will-win-the-game-of-thrones-1203)
It's really up to the discussion creator. Multis allow you to change the theses and add a "better" one, as happened in the drug legislation debate (I think).
Re 4) Do you remember nupedia, it didn't work. Wikipedia is not perfect, but the best resource we have to look up the "what". What we are building with the public site is a repository of the "why".
It doesn't require experts to write the reasoning of experts, which is what is happening, sometimes phds come in and post all the scientific papers pertaining to a position.
We agree that the hardest issue is actually moderation. When you have thousands of people chime in, as is the case with e.g. the UBI debate, you need a good system to filter for duplicates and trolls. That's why we implemented the suggestion system (https://support.kialo.com/hc/en-us/articles/115003791445-Sug...) and 5 different permission levels(https://support.kialo.com/hc/en-us/articles/115003877705-Par...).
This seems to work, as can be witnessed in the large debates. It's by no means perfect though and we are constantly improving it. The last bit we added was nested suggestions and suggestion commenting.
Again, thanks for your input. :)